Add YYLO Benchmark to AI & LLM Testing

[InsightFactoryAPP]

Hi maintainers! Thanks for maintaining this valuable list.

I'd like to propose adding YYLO Benchmark to the AI & LLM Testing section.

What is YYLO Benchmark?

YYLO Benchmark is a flexible, isolated evaluation layer for task prompts and project-owned Workflow Runner YAML. It normalizes every case into one attempt contract, runs each candidate in a private fresh-repository workspace (dedicated repo plus private home/temp/cache roots behind a selective filesystem sandbox), and evaluates the retained evidence with ordered deterministic and/or LLM judge profiles. Results chain receipts, manifests, and hash-verified terminals into immutable records, so regrade/rejudge consume retained evidence instead of re-running candidates.

• Repo: https://github.com/yylo-dev/yylo-benchmark (MIT)
• npm: @yylo/benchmark (yylo-benchmark CLI, also delegated by yy benchmark from the YYLO CLI)

Why this section?

The section's focus is exactly this noun: testing frameworks for AI systems (promptfoo, Tenro, nika, crilio). YYLO Benchmark occupies the agent-run side — the agent's executed attempt is the software under test, evaluated deterministically and by LLM judges with retained, tamper-evident evidence. It complements prompt-output regression tools like crilio and workflow snapshot engines like nika.

Contribution notes

• Single addition at the end of the ### AI & LLM Testing section, one-line format matching the existing entries, description ends with a full stop, no trailing whitespace.
• Searched the list first — no duplicate (no existing YYLO entry).

Disclosure

I'm part of the team that builds YYLO. This PR was prepared with an AI coding agent and reviewed by me before submission.